Although methadone maintenance has some distinct advantages compared to continued heroin use for pregnant heroin addicts, e.g., better prenatal care, bigger infants, it has the disadvantages of needing daily administration of a clinic to avoid withdrawal symptoms and diversion of take-home doses. Further, infants exposed to methadone in utero have a high incidence of moderate to severe abstinence symptoms and display various neurobehavioral deficiencies. Levo-alpha-acetylmethadol (LAAM), a long-acting opiate which is given three times a week, may be a better maintenance treatment. We hypothesize that LAAM will produce a less severe neonatal abstinence syndrome than methadone and facilitate patient compliance. At present there is no information regarding LAAM's disposition in pregnancy or its possible effects on the neonate. This project will determine the half life of elimination and excretion patterns of LAAM and its active metabolites during various periods of the pregnancy, at delivery, and after delivery in the neonate. Throughout the pregnancy, drug use questionnaires and urine toxicology screening will be done. At birth, abnormalities will be determined by physical and neurological exam. A Brazelton Behavioral Assessment will be performed to assess responsiveness. Signs and symptoms of an abstinence syndrome will be evaluated daily in the hospital and after discharge for 8 weeks. Rates of LAAM elimination will be correlated to incidence and severity of abstinence. Follow-up evaluations of the infants will be done repeatedly at 6, 9, 12, 15, 21 and 24 months. To determine neurobehavioral development, a neurological exam and the Bayley Scales of Mental Motor Development will be administered at six month intervals. Additionally, structured assessments of maternal functioning and maternal depressive symptoms will be performed. HOME assessment scale will also be administered. Multivariate analysis will be employed to determine relationships between LAAM and other drugs abused, perinatal and maternal psychosocial variables and neurodevelopmental outcome. Group comparisons of LAAM versus methadone will also be made. This study will determine the efficacy, disposition, and effects of pregnant women on LAAM versus methadone during pregnancy as well as the physical and neurodevelopment of the offspring. Using this information, further studies can be designed and intervention programs may be developed.